A Simulation-Based Variable Neighborhood Search Approach for Optimizing Cross-Training Policies

We study cross-training policies in a single multi-skill, multi-server repair facility with an inventory of ready-to-use spare parts. The repair facility has an inventory facility for different spare parts. If available, the failed spare parts are immediately replaced with new ones from inventory. Otherwise, the spare parts are backordered with penalty costs. This paper proposes a model to optimize skill assignments to minimize the system’s total cost, including servers, training, holding, and backorder costs. We develop a simulation-based variable neighborhood search approach, where we use discrete event simulation to evaluate backorder and holding costs under stochastic demand and service times. The simulation model is integrated with the optimization model to find the optimal skill distribution between servers. We tested the performance of our proposed framework by comparing its results with optimal solutions for small-size cases obtained using brute-force optimization. Also, we compared the performance of the proposed VNS algorithm to GA.